Should There Be a Compulsory License for Derivative Works?

By George Howard

Copyright law, which governs much of the music business, attempts to strike a balance between the rights of the creator, and the common good. That is, the law grants copyright holders a set of exclusive rights, but only for a limited time. At the end of that time (currently, seventy years after the death of the last-living author in the US), the work falls out of copyright, and can then be used by anyone; thereby contributing to the common good.

The Laws About Covering a Song

It’s not just this expiration of copyright that attempts to strike the balance between the rights of the individual and the common good. For instance, once an artist commercially releases a song, anyone who desires to can re-record the song (i.e. cover it), and release his or her version so long as the person who covers the song adheres to the rules. These rules, known as the compulsory or statutory license, define precisely what the artist who covers another’s work must do to avoid problems. Specifically, unless otherwise negotiated, the artist who creates the cover must pay the original writer (or her publisher) the maximum statutory rate for every reproduction of the cover. The rate is currently $.091 per reproductions (just under a dime) for songs under five minutes in length. Additionally, the person covering the song must account to the original author every month.

These requirements notwithstanding, the original author cannot stop another artist from covering his or her work. They are compelled to license their exclusive copyright to any other artist who wants to cover it, so long as the artist covering it abides by the rules emphasized in the copyright code. “Compelled” means that the original author of the song can’t say no.

Whether you agree or disagree with this approach, it should be fairly obvious that this compulsory license shows the law attempting to strike the delicate balance between giving the copyright holder his or her due rights, while still allowing others to create and benefit from their own covered works. The proof is somewhat in the pudding.

Covers are a pervasive and important part of the music business. Many artists benefit from the compulsory law by being able to easily (and without infringing) perform and record/reproduce/distribute their versions of another’s song. This often results in interesting new versions of works being created that, if not for the compulsory license regulations, would never have seen the light of day. Of course, because of these compulsory rules, the original author also benefits financially. Certainly, it may not be a perfect balance, but it seems to work.

How Derivatives Work

The fact that it does seem to work fairly well makes it all the more curious that such an approach has not been put in place for other elements of copyright. Specifically, the exclusive right to prepare derivative works. Derivatives are new works drawn (or “derived”) from another work. A simple example is a translation. Only the copyright holder of the original work can create derivatives. Therefore, if some other artist records a version of a work in a different language, this artist will be deemed to be infringing upon the original author’s exclusive right to prepare derivative works.

Of course, it goes way beyond translations. Samples are a form of derivative works. That is, if you pull a portion of someone else’s work and insert it into your own work, you are deemed to be creating a derivative of the work from which you sampled. You can’t do this without the consent of the rights holders of the original works: the copyright holder of the song itself (typically, the writer/publisher), and the copyright holder of the sound recording (typically, the label).

Unlike when you cover a song, and rely on the compulsory license regulations to both grant you the rights to cover the song in the first place, and to outline your responsibilities (amount you must pay, when you must pay, etc.), there is no such compulsory license when it comes to derivatives.

Any time an artist desires to sample a work (or “replay” it; that is, record his or her own version of a portion of a song rather than sampling it from a master recording) the artist must negotiate with the rights holders, and the rights holders can set whatever terms they want (or just deny the request all together).

Why is this?

It strikes me as odd that you can, without going through this expensive and arduous process, cover whatever song you want…in its entirety. But, if you want to sample five seconds of that same song, you can’t do so in any easy or cost-effective manner.

This discrepancy between there being a compulsory license for covers, but not for samples (or other derivatives) strikes me as not following the intent of copyright; it does not seem to strike the balance between the rights of the creator and the common good. It appears to err too far towards the rights of the creator, and leave the common good wanting.

There are likely arguments having to do with the rights of the author to control the context of his work at the root of this discrepancy. That is, the original author might not want her work sampled and added into a work that she finds repugnant. Fair enough. However, this “logic” doesn’t really hold up under scrutiny. There are plenty of covers that the original authors likely find repugnant, and yet they can’t do a thing about it.

A License to Consider

My proposal, therefore, is that we work towards a compulsory license for derivative works. Perhaps, it could be some sort of sliding scale. For a certain number of reproductions and streams, the sampler must pay the copyright holders a certain amount, and that amount increases when certain thresholds of reproductions or streams are met.

For instance, perhaps the artist being sampled might split 25% of the copyright of the work that samples their copyrighted works, so long as there are fewer than 10,000 reproductions and 100,000 streams. That amount could increase to 50% of the copyright from 10,001 reproductions to 100,000, and from 100,001 streams to 500,000. Maybe it increases from there. Importantly, neither the copyright holder of the composition nor the copyright holder of the sound recording can say no; they are compelled to allow for the derivative (sample) to take place.

Certainly, monitoring the number of reproductions and/or streams, and accounting accurately won’t be easy, but it’s not easy to monitor reproductions and/or streams for covers.

Creating a compulsory license for derivatives would not only satisfy the stated goal of copyright: the common good, but would also increase revenue-generating opportunities—for both the party doing the sampling, and the parties being sampled—at a time in the music business when we could certainly use some additional revenue being generated.

George Howard is the COO of Concert Vault, Daytrotter, and Paste Magazine. Mr. Howard is an Associate Professor of Management at Berklee College of Music. Follow George on Twitter.

  • 123mickey

    There are several problems concerning your idea. Two primary ones:

    A. One of the inherent rights in copyright law is the creator’s ability to discern the uses they allow for their material. The copyright law sets out to protect creators, in part by enabling them to make those exact choices rather than let their work be freely used, exploited and reproduced. The cover song compulsory mechanical license is an EXCEPTION to the law, and does not represent the overall spirit of intellectual property ownership legislation. Therefore, it’s anything but an obvious leap to apply the same compulsion towards sample usage.

    B. Unlike covers (or interpolations), sampling features actual recordings by other people. It could certainly be argued that those recordings, and the fact that they contain personal, recognizable voices and/or playing styles, are the inherent property of their owners and hold a more inextricable link to them than a vague “concept” of a song.

    Both points pertain to the fact that compulsory license for derivative works would infringe on personal, highly esteemed rights, and *much more so* than compulsory cover laws. That is why I believe such an amendment to the law would never see the light of day – and perhaps that’s not such a bad thing. For example, imagine what would happen if somebody took your song and replaced every single word with an F bomb. Or if somebody released a purposefully detuned version of your original.

    One last point, and perhaps the most practical one, is that there is simply NO easy automatic way to determine splits in a derivative work. It’s a matter of negotiations and leverage. Do you what’s Sting’s share on Puff Daddy’s “I’ll Be Missin’ You” (which samples “Every Breath You Take”)?100%.

    Why? Because he had the leverage.

    So… Yeah. I don’t see this working out.

    • thanks for your comment.

      re A, you’re wrong. there are no moral rights clauses in US (c) law. there are, of course, remedies avail, but only through tort. re b – i don’t follow.


      • MichaeIH

        re. A. – There are no moral rights clauses in the actual copyright law because they are covered elsewhere. When the US signed the Berne Convention (which does include Moral Rights), it stipulated that those provisions were addressed sufficiently by other statutes – and this stipulation was later affirmed by the Supreme Court. You’re welcome to refer to R. R. Kwall’s book “The Soul of Creativity” for more information. While their applicability and effectiveness is somewhat debatable, it’s a crass oversimplification to say that they simply don’t exist.

        Nevermind B, which would take a while to explain; this law would never pass because, unlike the clear cut case of cover songs (which only involve stat rate for the writers and a small regulated advance, depending on how you get the license), no statutory compensation rate will ever be successfully determined, either in regards to clearance fee or splits (both of which can be pretty ridiculous at times, depending on the licensor’s leverage and the licensee’s budget).

        Your idea may be idealistic (although I don’t share the ideal) – it’s just that at the same time it’s also extremely impractical.

  • Hi George,

    I was a student of yours in the Music Industry program at Loyola NO.

    I think you make some valid points about the way these two processes currently function. I am certainly a fan of simplifying the copyright laws as much as is feasibly possible (which isn’t much at all) and think that a Comp. License for derivatives would work at achieving that.

    I want to make sure I understand where you’re going with the idea of splitting the copyright of the derivative work, though. In your model, could the owners of the original copyright end up owning the copyright of the derivative, as well (depending on the negotiated terms, of course)? It’s almost like putting a Sunset Clause on what the derivative can make earn from the use of the original copyrighted material. As with anything legal this could get complicated, especially if the derivative samples several different works. Nevertheless, this is what attorneys are for! I guess all the original owners of sampled material could end up owning the derivative and splitting that equity. Like I said, complicated (and convoluted), but incredibly interesting.

    The more I think about it, the more I like it. It could be done.

    I would like to add that, in my opinion, there may a bit of vanity at play regarding the way the system works now. If my work gets covered and it bombs, then the cover-artist is the one that looks bad. I got paid for my work and they made a fool of themselves; no harm, no foul. However, if my work is sampled in a piece that bombs or is something I, on personal principle, don’t agree with then I’m associated with it; not just my sheet music, but my sound recording– my voice, my instrumentation. I can see where some may have a problem with that and why things function as they do. Just a thought.


    The law is there to protect the creator and his/her rights, but George is also right in that the law exists to create a balance. The law doesn’t exist to grant a monopoly to the artist, while removing all rights from the general populace. Some people would never make money off of their work if a monopoly existed in the creator’s favor and there may never be professional musicians if they weren’t protected. A balance must exist for such a symbiotic relationship.

    Also, the Comp. license is not the only exception. The First Sale Doctrine also protects the rights of the general populace. Once someone purchases a cd, they are free to do with it, physically, as they please (until they try to reproduce it). Now, this is concerning the physical good (e.g. cd, DVD, etc.), but the idea is the same. The creator must relinquish hold of the work at some point for there to be any kind of natural ebb and flow to the business. If you’re a musician and you want to sustain yourself and your family through music, then you are a part of the business.

    I enjoyed the article, George. Now, how would one go about actually implementing a change like this?

    • hey! great to hear from you.

      easiest way to think about your question/point about splitting the (c), is to think of it almost like a co-write. that is, once the derivative is created there are 2 or more people who may have a claim on some % of the (c) (in the same way as if you write a song with another person – he/she would have an interest – absent some agreement to the contrary – in the (c).


      • Right. I’m following you.

        I was also taking it further past the initial creation of the work. In the situation you created above, could the original copyright owner eventually own the derivative, as well? If the terms were set as you mentioned, where increased reproductions/streams increased the percentage of the original (c) owner’s equity in the new derivative?

        I’m trying to get at what could happen if the derivative were wildly successful, according to the parameters you laid down above.

        I also wanted to let you know, that your class was probably the most rewarding for me, intellectually and personally. It came at a good time, because I was falling out of love with the industry. Your passion for it came through and was a very strong influence. You also fed my brain, which is just as important! Thanks for that.

        • well, when the (c) is split, all rev from the (c) (mechanical income, perf, etc.) is divided amongst the (c) holders. once the (c) is divided – absent some agreement to the contrary (ie a license) – you can’t put it back together again, and so the orig holder would not be able to own the derivative unless the person who created the derivative makes a deal to do this.

  • Compulsory license works because it’s a flat rate per mechanical and publisher/writer retain 100% of song. That is cut and dry, easy to work with and automate. Some countries have laws regarding translations of songs — for example, translator gets 10% writing credit for a version of a foreign song in the local language. That’s more along the lines of what you’re thinking, but still cut and dry.

    Sample compulsory license *for publishing rights* could work if it was:

    % of length of song using sampled material x 9.1¢ (mechanical compulsory rate) = sample fee due per copy pressed/downloaded

    That percentage of your composition would also be owned by the sampled material’s publisher.

    If you use a sample of another song behind your entire song (Beck has done this), then the publisher of sampled song would own your entire composition, making it effectively a cover.

    For the sound recording copyright, you really have to negotiate directly. I’m guessing it’s usually a one-time fee there. If you’re sampling a #1 hit song, you pay thru the nose. If it’s obscure music, you pay very little. There’s no way you should be able to sample Sting’s voice and Andy Summers playing a riff and pay the same as you would to sample the punk band down the street. It has to be arbitrary.

  • Tony Kosinec

    Sampling a recording involves not just intellectual property, but something closer to real property. There are people playing those sounds and new masters embodying their performances are new “sessions” for which they are due payment by the owner of the production copyright when they were recorded under a musician union contract. – These individuals appear as artists in the sampled production. – and it’s not just the idea being used, it’s the (costly) realization of the idea. To my thinking using a recording is like renting a room in someone’s house, there shouldn’t be squatters rights.

  • SJ

    Interesting….I completely follow your logic…however, I thought that the initial intent of derivatives under copyright law applied to the rights of the original copyright holders “only”, to create derivatives of their own work, if they so choose, and that this in itself would also be it’s own copyright under separate registration. Additionally, co-copyright holders are considered equal parties with 100% ownership (unless stipulated by separate agreement); however, when doing a derivative of one’s own work…permission from the other copyright holders is not necessary. However, all parties cannot be divested of their ownership and retain full rights of even derivative works. (That according to the U.S. Copyright Office)
    If what you’re saying is correct…that would be a cost effective an ideal situation for all parties. Just a clarification as to where I’m confused….if someone (not an existing copyright holder) were to follow the course you recommended, the “covering artist” would only need to get permission from one of the copyright holders to move forward with a derivative work? How would this work? The law is clear that copyright holders cannot be divested or diluted (like shares) from their own work. In this particular case, even if permission were given (not sure how that process would work), the “covering artist” would see absolutely no $.? Am I correct?

  • Patrick

    I’ll have to disagree. Though I might not like someone’s cover of my tune, their rendition does have maintain the integrity of the tune, and if it doesn’t, it’s a derivative, and if it’s a derivate, I, as composer, I would want the artist to ask my permission to see if I like what the artist wanted to do, and we negotiate terms if I accept it. This idea that so and so wants to sample two bars of my hook, for which my song got its fame, and they are riding on the coatails of my tune for their new tune for merely a compulsory license fee does not cut it, with me. Therefore, the status quo is correct, in my view.

    • Patrick

      I’ve thought about this even more. If someone wants to use part of my tune, say they want to sample a few bars and make a new tune with it, the new tune is a derivative with me as co-author. You can’t force someone to co-author a song, that would be tyranny, and so compulsory licenses for derivatives simply will never happen for this reason. Tyranny is not in the public interest, period.

  • Erik

    It should be an opt in or opt out license which is available when cutting a deal with a label and publisher. The fact is that people are going to make these remixes or mashups anyway. It could be an additional revenue stream which is currently being viewed as a problem.

  • Patrick

    Title 17 states that the right to create derivative works ( or license to others to do so) is that of the composer. Your proposal wants to take away this right? NO WAY. This is wrong, plain wrong.

  • VladanMovies

    Point is not in liking, or not liking someones work, but in money. If someone makes derivative without mentioning original, original author can not get any royalties, because he can get them only for licensed work, and this derivative he can not license because it is original work by someone else. In short, it’s about the title of the song. Author registered it by title. Derivative bears completely different title, so there’s no way author of original material, that was used in derivative, can licence it, less the current way of negotiation on “per case basis”.